Autism is one of most heritable neurodevelopmental disorders with genetic basis well established through family and twin epidemiological studies. Genome-wide association studies have not found strong evidence for the contribution of common variants, and linkage studies indicate the presence of multiple loci, each of which contributes negligibly to the genetic variants on a population level. The goal of this proposal is to apply novel analytic approach to identify families from existing UW-ACE and NIH collections where inheritance is most parsimonious with single gene transmission, and to apply novel genomic technologies to identify the genes that are responsible. To identify families where the likelihood of dominant and recessive inheritance is increased we will apply identity-by-descent an extent-of-homozygosity analyses to the existing genotype data. Next, we will capitalize on newly available methods for a whole genome evaluation of protein coding regions for sequence and copy number variants that might be causal. Lastly, the candidate genes identified will be evaluated for association with autism in a large case control study and functional analyses of candidate genes with the highest evidence for causality will be initiated. To facilitate validation of our findings we will deposit genotype information in NIMH data-base. The understanding of genetics of autism will facilitate early interventions by enabling presymptomatic diagnosis, implicate additional biological pathways involved in autism and increase the number of targets for causative treatments.